LightCurves.md

Previous: Binary lenses · Documentation home · Next: Critical curves and caustics

Light Curve Functions

VBMicrolensing correspondence: LightCurves.md. The binary- and triple-lens values and example order are preserved.

Binary Lens light curve

The values below are used unchanged throughout the calculation.

import numpy as np
import matplotlib.pyplot as plt
import lcbinint

s = 0.9
q = 0.1
u0 = 0.0
alpha = 1.0
rho = 0.01
tE = 30.0
t0 = 7500

params = {
    "s": s, "q": q, "u0": u0, "alpha": alpha,
    "rho": rho, "tE": tE, "t0": t0,
}
t = np.linspace(t0 - tE, t0 + tE, 300)

curve = lcbinint.LightCurve(options=lcbinint.Options(tol=1e-3, reltol=1e-3))
magnifications = curve(t, params)
trajectory = curve.source_trajectory(t, params)

Plot only the light curve:

plt.figure(figsize=(3.8, 2.55))
plt.plot(t, magnifications)
plt.xlabel("Time")
plt.ylabel("Magnification")
plt.show()

Calculate and plot the caustics and source trajectory separately:

caustics = curve.caustics(params)

plt.figure(figsize=(2.8, 2.8))
for x, y in zip(caustics.x, caustics.y):
    plt.plot(x, y, color="tab:red", lw=1.1)
plt.plot(trajectory.x, trajectory.y, color="tab:blue")
plt.xlabel("X")
plt.ylabel("Y")
plt.axis("equal")
plt.show()

Warm up a fixed binary light curve

Repeated native evaluations of exactly the same binary-lens parameter vector and epoch array can retain a specialized execution plan:

report = curve.warmup(t, params)
magnifications = curve(t, params)  # automatically uses the retained plan

The report is diagnostic and does not need to be passed back to curve. For each epoch it records the retained method, measured resolution, calibration status, reference value, error budget, and Cartesian/polar timing. Point, hexadecapole, and source-plane decisions are retained from ordinary automatic evaluation. Inverse-ray epochs measure qualifying Cartesian and polar grids and retain the faster one.

Plans are exact-keyed. The times, parameters, tolerances, numerical options, model configuration, and limb darkening must match the warm-up call. Any mismatch uses ordinary automatic evaluation; an epoch that could not be calibrated also falls back independently. Use curve.clear_warmup() to discard the plan. Warm-up currently supports native single-source binary-lens curves, not JAX, binary-source curves, or triple lenses.

The production resolution law used as the search hint and its supported tolerance domain are documented in Calibration of automatic finite-source resolution.

Triple Lens light curve

s = 0.9
q = 0.1
u0 = 0.0
alpha = 1.0
rho = 0.01
tE = 30.0
t0 = 7500
s13 = 1.5
q3 = 0.003
psi = 1.0

params = {
    "s": s, "q": q, "u0": u0, "alpha": alpha,
    "rho": rho, "tE": tE, "t0": t0,
    "sep2": s13, "q2": q3, "ang": psi,
}
t = np.linspace(t0 - tE, t0 + tE, 300)

curve = lcbinint.LightCurve(
    lens="triple",
    options=lcbinint.Options(tol=1e-3, reltol=1e-3),
)
magnifications = curve(t, params)
trajectory = curve.source_trajectory(t, params)

Plot only the light curve:

plt.figure(figsize=(3.8, 2.55))
plt.plot(t, magnifications)
plt.xlabel("Time")
plt.ylabel("Magnification")
plt.show()

Calculate and plot the caustics and source trajectory separately:

caustics = curve.caustics(params)

plt.figure(figsize=(2.8, 2.8))
for x, y in zip(caustics.x, caustics.y):
    plt.plot(x, y, color="tab:red", lw=1.1)
plt.plot(trajectory.x, trajectory.y, color="tab:blue")
plt.xlabel("X")
plt.ylabel("Y")
plt.axis("equal")
plt.show()

Previous: Binary lenses · Documentation home · Next: Critical curves and caustics